Beta · Last modified February 25, 2010
The jury is out on tag clouds. Are they truly useful or just gee whiz cool to look at?
The common pattern is to present tags alphabetically and then enlarge the tags proportionally based on popularity. This presentation is popular but not always easily understood by users and can take up a lot of real estate in the interface.
User wants to know what all the tags are associated with an object, a site, a person and which ones are used most often.
The usefulness of a tag cloud is still under debate. They look neat and make a nice graphical element, but the nuances of the sizing differences are often lost on the average user. If the item or the site has a lot of tags, then the cloud becomes less usable than a site or object with fewer tags. Sites such as Flickr, which has thousands of tags, has organized its presentation of tags by dividing them into their "Explore" area for public tags and "Your Tags" which presents a subset of the most used tags for a person's items. In both cases, they present a small subset of the total.
Additionally, some sites display the different sized tags to represent some editorial emphasis or other data besides popularity or implied ranking. The lack of a standard meaning behind the tag cloud presentation makes it difficult for end users to know what to expect or to be able to predict the inherent meaning in the visualization of the tag cloud.
Using the tag cloud presentation, popularized by Flickr, is a good way to show users a visual representation of the concepts popular to a person's objects or the site. A tag cloud encourages browsing and exploration in an alternative way from the site's standard navigation and supports serendipitous discovery.